Tone in global-issue passages is rarely extreme; it's typically measured, analytical, or cautiously optimistic. Train yourself to catch rhetorical signals like purpose verbs (proposes, evaluates, critiques), hedging language (may, suggests, likely), and contrast markers (however, yet, nevertheless). When identifying tone, pick the most moderate accurate label supported by the evidence rather than your prior beliefs.

Evidence questions typically come in linked pairs: one asks for the best content answer, and the next asks which lines support it. Solve these as a unit by predicting first, then jumping to the evidence question to find which option actually proves your predicted idea. If no evidence choice backs your content answer, you likely misread and should adjust your content selection.

Start by reading all labels including title, axes, units, categories, and time frames. Next, identify what specific comparison the question asks for. Finally, confirm your answer by reading only the exact data needed. Be careful to distinguish percent change from percentage-point change, and avoid claims that reference categories or time spans outside what the graph displays.

One major pitfall is treating percentages as absolute counts without knowing the base size. Another is confusing a trend (rising line) with the actual level (the line may still be below others). Also avoid reading beyond the axis range or assuming causation from correlation—look for explicit causal language before claiming one factor caused another.

First replace the word with your own plain paraphrase that fits the sentence, then match to the closest answer choice. For domain jargon, identify the term's role (variable, outcome, mechanism, constraint) and substitute an everyday synonym that preserves the logic. When two answers seem plausible, choose the one most constrained by the sentence context rather than a broader dictionary definition.